Series.replace cannot use dict-like to_replace and non-None value

Asked Nov 13 '20 at 09:44

Active Nov 13 '20 at 09:44

Viewed 3,995 times

I have a DataFrame with the fields as shown. Dataframe fields

I am trying to fill the null values with the median of the column Loan_Amount_Term with the below code: (The column has 14 null values)

hd['Loan_Amount_Term'] = hd['Loan_Amount_Term'].replace(np.isnan(hd['Loan_Amount_Term']),hd['Loan_Amount_Term'].median())

But I am getting the following error: Series.replace cannot use dict-like to_replace and non-None value

I want to know why I cannot impute the null values with median in this way.

Thanks in advance.

asked Nov 13 '20 at 09:44

Equalizer

change `.mean` to `.median` and use [this](https://stackoverflow.com/a/33178896) – jezrael Nov 13 '20 at 09:45
@jezrael this is replacing the entire **Loan_amount_term** column to **None**. – Equalizer Nov 13 '20 at 09:53
then use `df.Loan_amount_term = df.Loan_amount_term.replace('None', np.nan)` before this solution – jezrael Nov 13 '20 at 09:54
1

It works fine if I remove **inplace = True** argument. Thanks @jezrael But I want to know why **replace** is not a good fit here. – Equalizer Nov 13 '20 at 10:04

0 Answers0